AI assistants

PortMem in your AI assistant.

Connect PortMem to Claude, ChatGPT, Gemini, Le Chat, Perplexity, Muse, Microsoft 365 Copilot or Slackbot and the assistant can work from your own source documents and check a draft against them before you rely on it. Every verdict comes with the passage that decides it. The decisions that matter stay with you, in the PortMem console.

Connector address
https://mcp.portmem.com/mcp

Connect it

Claude. Open Customize, then Connectors, then Add and Add custom connector. Paste the address and choose Add. Claude opens PortMem: sign in and approve.

ChatGPT. PortMem is in ChatGPT's plugin directory. Open Plugins, search for PortMem (by PortMan LLC) and add it. ChatGPT opens PortMem: sign in and approve. You can also add it as a custom MCP server: open Plugins, then Add and Create custom MCP server, paste the address with OAuth sign-in and create it.

Gemini. On a computer, go to gemini.google.com. At the bottom, open Settings, then Connected Apps (if you do not see it, open Personal Intelligence first). Under Custom apps, paste the address in the field to add a custom app and choose Next. Follow the steps on screen: Gemini opens PortMem, where you sign in and approve. To use it, type @ in a chat and choose PortMem. Once connected, it also works in the Gemini mobile app and in a Gemini Task (Google renamed Spark to Task).

Google offers custom apps to people 18 or over in the US, signed in with a personal Google account (not a work or school account), in English, with Keep Activity on. Tasks also need a Google AI Pro or Ultra plan.

Perplexity. Open Connectors, then Custom connector, then Add MCP connector. Name it PortMem, paste the address, confirm the notice and choose Add. Then open PortMem in the list and choose Add connector: Perplexity opens PortMem, sign in and approve. Custom connectors need a Perplexity Pro, Max or Enterprise plan.

Le Chat. Open Connectors, then Add Connector, then the Custom MCP Connector tab. Name it portmem, paste the address and choose Connect. Le Chat opens PortMem: sign in and approve. Adding connectors is for the account's administrator.

Slackbot. Slackbot's AI, which is what uses apps, comes with Slack's Business+ and Enterprise+ plans. A Slack admin first adds the PortMem app to the Slack workspace (see PortMem in Slack). Then each person opens Slackbot, opens its sidebar, then Apps, then the Approved by admin tab, and chooses Connect next to portmem: Slack opens PortMem, sign in and approve. Slack needs a full connection, so leave Read only unticked. Slackbot asks before it uses a PortMem tool. If your PortMem workspace is a company workspace, a PortMem admin first allows Slackbot in Settings, Assistants.

Microsoft 365 Copilot. A Microsoft 365 administrator adds PortMem as a connector, and Copilot can then search and read each person's PortMem documents and reviews (Microsoft's connectors are read only). Register once for a client ID and secret:

curl -s https://mcp.portmem.com/register -H 'content-type: application/json' -d '{"client_name":"Microsoft 365 Copilot","redirect_uris":["https://teams.microsoft.com/api/platform/v1.0/oAuthRedirect"],"grant_types":["authorization_code","refresh_token"],"response_types":["code"],"token_endpoint_auth_method":"client_secret_post"}'

In the Teams Developer Portal, open Tools, then OAuth Client Registration, then New OAuth connection. Enter the client ID and secret from that reply, the authorization endpoint https://mcp.portmem.com/authorize, the token and refresh endpoints https://mcp.portmem.com/token, the scopes portmem.read offline_access, and enable PKCE. Copy the registration ID it shows. Then in the Microsoft 365 admin center open Copilot, then Connectors, then Gallery, then Create a new connector and Connect to MCP server: name it PortMem, enter the base URL https://mcp.portmem.com/mcp and the registration ID, and roll it out to a test group first. Each person signs in to PortMem the first time Copilot uses it. Keep the secret private.

Gemini Enterprise. An administrator adds PortMem as a custom MCP server (Google's steps for the Standard, Plus, Pay-as-you-go and Frontline editions). Get a client ID and secret for your organization by registering once:

curl -s https://mcp.portmem.com/register -H 'content-type: application/json' -d '{"client_name":"Gemini Enterprise","redirect_uris":["https://vertexaisearch.cloud.google.com/oauth-redirect"],"grant_types":["authorization_code","refresh_token"],"response_types":["code"],"token_endpoint_auth_method":"client_secret_post"}'

Then, in the Google Cloud console, open Gemini Enterprise, then Data stores, then Create data store, and choose Custom MCP Server. Under authentication choose OAuth 2.0, and enter the server URL https://mcp.portmem.com/mcp, the authorization URL https://mcp.portmem.com/authorize, the token URL https://mcp.portmem.com/token, the client ID and secret from that reply, the scopes portmem offline_access (or portmem.read offline_access for read only), and tick Enable PKCE Support. Choose Verify Auth, then Continue, give the connector a name and choose Create. Google turns every action off at first: open Actions, choose Reload custom actions, select PortMem's actions and choose Enable actions. If your organization limits outbound traffic (an egress policy or VPC Service Controls), allow the domain mcp.portmem.com and the custom_mcp data source. Keep the secret private.

In a PortMem company workspace, AI assistants are off until a workspace admin turns them on in the console under Settings, Assistants.

You need a PortMem account. Each person connects with their own account and the assistant only sees that person's workspace. You can see and disconnect your assistants at any time in the console under Settings, Assistants; access stops at once.

Claude auto-connect, for administrators

People in a company on Claude Team or Enterprise can use PortMem in Claude without signing in to PortMem or approving it. Claude vouches that each person belongs to your Claude organization (Claude's managed authorization, Claude-issued), and PortMem signs in the member of your company workspace with the same verified email address.

Claude shows this option for PortMem once Anthropic has marked PortMem's directory listing as supporting it. Until then, people connect as described above.

What you need. A PortMem company workspace where you are a workspace admin, and a Claude Team or Enterprise organization where you are an Owner or Primary Owner. Claude does not offer Claude-issued sign-in in an organization that signs in to Claude through Okta. Each person's Claude email must be the address they use in PortMem, that address must be verified, and they must be a member of the workspace (Settings, Team). Someone new signs in to PortMem once first, which creates their account.

1. In the PortMem console, open Settings, Assistants. Under Allowed assistants, choose Only these and tick Claude, or choose All.

2. In Claude, open Organization settings, Connectors. Find PortMem (add it from the directory if it is not there) and choose Set up auto-connect. Choose Claude-issued as the identity source.

3. Still in Claude, enter your workspace's tenant ID. The PortMem console shows it under Settings, Assistants, Claude auto-connect; it starts with o-. Then copy the Issuer URL Claude shows. It starts with https://identity.anthropic.com/issuers/.

4. In the PortMem console, under Claude auto-connect, paste the Issuer URL and choose Turn on.

5. Back in Claude, run the connection test, then choose who gets the connector.

What happens then. Each person is connected to PortMem the first time Claude needs it, with the same tools as a connection they approved themselves. Settings, Assistants lists these connections as set up by your company. Their access lasts an hour at a time, and Claude renews it by vouching for the person again. Someone who is not a member of the workspace, or was removed from it, is refused. When you remove someone from your Claude organization, Claude stops vouching for them; remove them from the PortMem workspace as well.

Turning it off. Choose Turn off under Claude auto-connect. Every connection it made ends at once. Each change to this setting, and each connection it makes, is recorded in the workspace's signed ledger.

What the assistant can do

List your documents. Your own documents, those shared with you, and public reference documents.

Survey a document. The parties, key dates, amounts and material provisions of one document, read from the whole text. The first survey of a long document can take a few minutes; the assistant is told to search meanwhile, and the result is kept for next time.

Search. The passages that answer a question, with the document each one comes from.

Add a document. Text you give the assistant becomes a private document in your workspace. The assistant only adds new documents. Replacing a document you already have is done in the console.

Check a draft. A review of a memo, summary or report against your documents. Each claim comes back supported, contradicted or unsupported with its evidence, along with the material provisions the draft leaves out. Long reviews run in the background and the assistant checks on them.

Open past reviews. Your earlier reviews and their records, with a link to each in the console.

What stays with you

An assistant cannot accept a flagged claim, publish a record or change who can see a document. Those are your decisions, made in the PortMem console, where they become part of the signed record. When one is needed, the assistant gives you the link.

Read only. When you approve an assistant you can tick Read only. It can then read and search your documents and open your reviews, but cannot add a document or start a review. Settings, Assistants shows which connections are read only.

Tool reference

Every tool works only inside the signed-in person's own workspace. No tool takes a payment, sends a message, publishes, shares or deletes anything. Read tools change nothing. The two write tools only add something new, a document or a review record, and neither is a sensitive write. A read-only connection has the read tools only.

list_sources (read). No inputs. The names of the documents the person can use: their own, those shared with them, and public reference documents.

search (read). Inputs: query; optional source or sources (document names), max_results (default 8), expand (adds the surrounding text). Returns matching passages, each with the document it comes from.

survey_source (read). Inputs: source; optional refresh. Returns the parties, key dates, amounts and material provisions of one document, read from the whole text. The first survey of a long document can answer that it is still preparing; the result is then kept, so asking again is instant. The document itself is not changed.

upload_source (write). Inputs: source_name, text. Adds the text as a new private document. If a document with that name already exists nothing changes and the tool says so; sending the same text again changes nothing.

start_review (write). Inputs: document (the draft); optional source, sources or source_groups to check against, max_claims (default 30). Starts a review in the background and returns a job_id. When it finishes, a new review record is added to the person's ledger. It accepts, publishes and shares nothing.

review_status (read). Input: job_id. Returns the status (queued, running, done, error or cancelled) and progress; once done, every claim with its verdict and evidence, the provisions the draft leaves out, and the console link for the record.

past_reviews (read). Optional limit (default 25). The person's earlier review records, newest first.

get_decision (read). Input: request_id. One review record: its claims, verdicts, whether it is signed, and its console link.

list_signoffs (read). Input: request_id. The sign-offs people recorded on that record in the console.

usage (read). No inputs. The workspace's queries and estimated cost this month.

capabilities (read). No inputs. The retrieval features this deployment has switched on.

Errors and limits. A tool that cannot do what was asked returns a result with an error message saying why, for example an unknown document name, an existing document name, or a review that is still running. An expired or disconnected sign-in gets HTTP 401 and the assistant signs in again. A review can take a few minutes, so the assistant polls review_status. Usage counts toward the workspace like any other use of PortMem.

Ready-made prompts

PortMem offers three in the assistant's menu: Check a draft against my documents, Key facts of a document, and Find something in my documents. Or just ask, for example:

"Check this draft against my credit agreement and tell me what it gets wrong."
"What are the key dates and amounts in the bond terms?"
"Find what my documents say about events of default."

Your data

The assistant receives only what the PortMem tools it calls return, such as search passages, survey facts and review results; your assistant's provider then handles that under your agreement with them. Government ID numbers (CPF, CNPJ, SSN) in those results are masked. PortMem stores the connection (the assistant's name and when you connected) and its sign-in tokens only as one-way hashes. Usage through an assistant counts toward your workspace like any other. The privacy policy has the details.

Technical details

A remote MCP server over Streamable HTTP at https://mcp.portmem.com/mcp, stateless, serving MCP protocol 2026-07-28 and earlier versions. Sign-in is OAuth 2.1 with PKCE, with client ID metadata documents or dynamic client registration. Claude auto-connect uses the JWT bearer grant (RFC 7523) with identity assertion authorization grants (ID-JAG) from Claude-issued issuers. Every tool carries read-only, destructive and open-world hints.

Help

Write to info@portmem.com. If a review looks wrong, send its record identifier (it looks like PM-WSVR-45R4-KX3M) and we can find exactly what ran without asking for your documents. More on the support page.